Terry's DRSSTC - New H-Bridge :-)

["Tesla list" <tesla@xxxxxxxxxx>]

Original poster: Terry Fritz <teslalist@xxxxxxxxxxxxxxxxxxxxxxx>

Hi All,

Well I got all the new H-bridge parts at noon today right on time :-)

Just about an hour to put it together and another to install it.

It is just like the old one but with bigger traces where the last one 
broke, no MOV, and the new 10nF ceramic buss caps.  I forgot to take 
pictures...   But it "looks" just the same...  I have all the covers on the 
box "these days" to contain explosions ;-))

The schematic is here:

http://drsstc.com/~terrell/schematics/NewH-Bridge.gif

The old MOV across the output was really not needed since I have others on 
the buss side of the bridge* and the gate drives have TVS...  The MOV acted 
much like a 450pF cap and models suggest that "any" capacitance directly 
across the bridge is a very bad thing!  I never could get a model to 
simulate what blew the last bridge.  The models all crashed due to giant 
di/dt errors.  So I guess the models just blew up too ;-))  So the "rule" 
is no caps directly across the H-bridge!!

I powered the buss first with five 9V batteries (~50 volts) and found that 
I had the feedback CT reversed since I goofed with it.  No big deal to 
switch the wires...  This confirmed that all was working well with a very 
"weak" power supply that would not hurt anything if something was messed up.

At full power to the buss, I did note it now has a significant "sound" as 
the bridge switches.  Magneto-constriction I think.  I did not get my ears 
next to it since another explosion would be "bad" with an ear near it.  But 
something is far noisier now.  Might just be a loose heat sink or two parts 
clattering too.  No big deal for now...

So I cranked it up into a bare primary circuit with about 3 ohms of resistance:

http://drsstc.com/~terrell/pictures/NewBridge-01.gif

Note that the voltage driving square wave is super clean!!  Almost "too" 
clean...  It started out clean with the last bridge too, but then got 
noisy...  Hmmmmmmm  ?;-|  I wonder if there was some damage to the old 
bridge I did not detect...  Gee whiz, I checked and double checked 
everything...  Think more on that...  But clean is good for now ;-)

Notice the 1.4MHz unloaded ring down:

http://drsstc.com/~terrell/pictures/NewBridge-02.gif

Just a resonance "somewhere".  Not sure at all where.  Does no harm, but I 
should probably "find it"...

The current limit circuit still trips too soon.  Probably noise on the 
protection inputs due to these super fast switching times!!!!

http://drsstc.com/~terrell/pictures/NewBridge-03.gif

http://drsstc.com/~terrell/pictures/NewBridge-04.gif

I did add a grounded shield to the CTs during the down time and "special" 
caps to the protection circuit:

http://drsstc.com/~terrell/pictures/CTshield.JPG

http://drsstc.com/~terrell/pictures/ProtCardCaps.JPG

These are just the output float pull down times, but even with the probe's 
100MHz limits, it is still hitting "gulp" 37500v/uS!!!!!!  Nice the that 
rise and fall waveforms are almost mirror image...

Right now the output in these super short time frames is just like an open 
load and there is nothing to push the output voltage back down once the 
IGBTs go all open during the transition.  I was thinking of adding some 
resistance across the output to see if that would help, but stopped 
remembering that flippantly putting things across the output is how I blew 
the crappojebbies out of the last H-bridge ;-))

So it is all up and working again.

Few question as to noise and such, but these are pure "science" problems 
that can be solved just by thinking *;-)  The performance is almost "way 
TOO good"...

I see no real problem in running the switch times SUPER fast!  Have to be 
very carful not to cross-conduct, but in this low duty cycle application 
the IGBT drives and IGBTs seem to love it ;-))  Goodness knows that the 
power lost to IGBT heat due to switching is about zero with 4nS transitions 
;-))  But the spike noise is pretty bad too...  Might have to consider 
massive filtering on the over current, or just command the *&^& thing to 
the "disable" state ;-))  I guess Steve's is working well so I will have to 
study his solution more...

Cheers,

	Terry

*There is a potential "big problem" if a DRSSTC driver were to blow 
open.  The say 20KV voltage on the Tesla coil's primary caps could find 
it's way inside the DRSSTC "box" and do "massive" damage!!!  Simple MOVs in 
the right places will "catch" these voltages in the case of say an H-bridge 
nuke and just short them out.